Designed and optimized novel protein and antibody-fusion constructs using computational modeling, structure-based design, and AI/ML generative models Performed in silico developability and liability assessments (solubility, stability, aggregation risk, immunogenicity, binding affinity) using advanced predictive and simulation tools Prioritized and selected lead candidates through multi-parameter optimization, integrating computational scores with project goals and risk–benefit considerations Coordinated expression, purification, and analytical characterization of new designs with CRO partners; prepared technical scopes of work and experimental plans Authored R&D reports, experimental protocols, and data summaries to guide functional assay design, inform portfolio decisions, and align cross-functional project teams

Developed predictive ML models for clinical outcomes in pediatric asthma integrating transcriptomics, clinical data, and microbiome features in a high-dimensional environment Validated pipelines with real-world data; applied SHAP to identify top biomarkers and risk factors Conducted imputation (MICE), class balancing, feature selection, model tuning (GridSearchCV, Optuna), model validation and visualization (PCA, volcano plots, clustering) Integrated multi-omics data for risk stratification using interpretable ML

Contributed to the Development text-to-protein generative AI models (MP4), Designed and evaluated de novo binders (Fc domain, GLP-1/GIP/GCGR peptides) using AF2, DiffDock, PyRosetta, BindCraft, ScanNet, and structure-based modeling with PDB-derived interaction analysis tools Built and analyzed prompt matrix for pLLM 310.ai Copilot, identified errors, debugged protein generation flows, contributed to the Documentation, fixing issues of pLLM 310.ai Copilot (GitHub)

A/B testing, Using APIs to Leverage Crowdtesting

Ensure the accuracy, completeness, and reliability of data used for training ML models. Assist in the development and training of ML models by preparing datasets, running training sessions, and evaluating model performance through various metrics. Work on validating the models against real-world scenarios to ensure they meet the desired accuracy and efficiency standards.

Tech stack: SQL, R, Python, BioPython, Databricks, Hadoop, Apache Spark, Kafka, Flask Machine learning Models: linear regression, logistic regression, Naive Bayes, decision trees, random forest, AdaBoost, XGBoost, k-NN, Gaussian Mixture Models, CNN, RNN, SVM Pytorch/TensorFlow/Keras Python packages: Pandas, NumPy, PySpark, Seaborn, Scipy, statsmodels, Scikit-learn Data Visualization: Matplotlib, Tableau, PowerBI, Google Sheet Structural Modelling: Chimera & PyMOL, Image data analysis with AI & Python, Molecular Dynamics: VMD & NAMD, OpenMM. Portfolio: https://github.com/matusoff https://app.usebraintrust.com/talent/329901

Development of small molecule drugs using chemoinformatics and docking (docking, chemoinformatics), Python programming for molecular dynamics, structural analysis, chemoinformatics)

Maintained accurate records of research findings and provided statistical analysis of data results; Use Excel and statistical software to analyze data and determine data-set correlations

Research conception/Design/Experiments to address defined scientific problems. Functionality of cells, tissues, and biomolecules in various pathophysiological conditions. GraphPad Prism/Excel/Data analytics/Statistical analysis. Present at conferences/publish research papers in peer-reviewed journals/write grant proposals. Leverage interpersonal and communication skills to mentor Ph.D./Master/undergraduate & summer students.